Fibre optic communication systems have gained widespread acceptance over the past few decades. With the advent of optical fibre, communication signals are transmitted as light propagating along a fibre supporting total internal reflection of the light propagating therein. Many communication systems rely on optical communications because they are less susceptible to noise induced by external sources and are capable of supporting very high-speed carrier signals and increased bandwidth. Unfortunately, optical fibre components are bulky and often require hand assembly resulting in low yields and high costs. One modern approach to automated manufacturing in the field of communications is integration. Integrated electronic circuits (ICs) are well known and their widespread use in every field is a clear indication of their cost effectiveness and robustness. A similar approach to optical communication components could prove helpful.
Unfortunately, integrated optical devices are generally quite lossy resulting in what is often unacceptable performance. In order to compensate for the performance of a lossy device, one approach is to use external optical amplifiers to amplify the light provided to the lossy component. Unfortunately, because of non-linearities in the optical amplification gain profile and in optical losses of an integrated component, results vary when changing either of these two devices in an optical system and generally, this approach is not preferable.
For example, a typical erbium doped fibre EDF produces differing gains for different wavelength channels when it is pumped. Using this technology in a fibre amplifier requires some means of ensuring that the different wavelength channels each receive the same amount of optical amplification. To compensate for the differing gains, a gain flattening filter is introduced to the amplifier assembly. The response of this filter assists in flattening amplification of the EDF. Unfortunately, the filter and the EDF are never perfectly matched. Consequently, the problem is not eliminated but merely reduced. Additionally, the gain flattening filter is a passive device with optical attenuation. Consequently, all the optical signals will loose some optical intensity when this component is used.
Of course, all of the above does not even begin to address manufacturing variations and tolerances wherein a filter response is within a range of acceptable filter responses and not at the exact, design specified, response. Thus, with an EDFA that varies from the design specified amplification response and a filter that similarly varies, the result is often the sum of the two variations.
It would be advantageous to produce an integrated device with minimal or no loss. Additionally, it would be advantageous to produce an integrated device providing dynamic gain equalization through amplification. Such a dynamic gain equalizer will act to replace a conventional dynamic gain equalizer combined with an amplifier providing a smaller and less expensive optical component.